1

Height and body mass index as modifiers of breast cancer risk in 22,588 carriers of BRCA1

or BRCA2 mutations: A Mendelian randomization study

Frank Qian1, Shengfeng Wang2, Jonathan Mitchell3, Lesley McGuffog4, Daniel Barrowdale4, Goska Leslie4, Jan C. Oosterwijk5, Wendy K. Chung6, D. Gareth Evans7, Christoph Engel8, 9, Karin Kast10, Cora M. Aalfs11, Muriel A. Adank12, Julian Adlard13, Bjarni A. Agnarsson14, 15, Kristiina Aittomäki16, Elisa Alducci17, Irene L. Andrulis18, 19, Banu K. Arun20, Margreet G.E.M. Ausems21, Jacopo Azzollini22, Emmanuelle Barouk-Simonet23, Julian Barwell24, Muriel Belotti25, Javier Benitez26, 27, Andreas Berger28, Ake Borg29, Angela R. Bradbury30, Joan Brunet31, Saundra S. Buys32, Trinidad Caldes33, Maria A. Caligo34, Ian Campbell35, 36, Sandrine M. Caputo25, Jocelyne Chiquette37, Kathleen B.M. Claes38, J. Margriet Collée39, Fergus J. Couch40, Isabelle Coupier41, Mary B. Daly42, Rosemarie Davidson43, Orland Diez44, Susan M. Domchek30, Alan Donaldson45, Cecilia M. Dorfling46, Ros Eeles47, Lidia Feliubadaló48, Lenka Foretova49, Jeffrey Fowler50, Eitan Friedman51, 52, Debra Frost4, Patricia A. Ganz53, Judy Garber54, Vanesa Garcia-Barberan33, Gord Glendon18, Andrew K. Godwin55, Encarna B. Gómez Garcia56, Jacek Gronwald57, Eric Hahnen58-60, Ute Hamann61, Alex Henderson62, Carolyn B. Hendricks63, John L. Hopper64, Peter J. Hulick65, 66, Evgeny N. Imyanitov67, Claudine Isaacs68, Louise Izatt69, Ángel Izquierdo31, Anna Jakubowska57, Katarzyna Kaczmarek57, Eunyoung Kang70, Beth Y. Karlan71, Carolien M. Kets72, Sung-Won Kim73, Zisun Kim74, Ava Kwong75-77, Yael Laitman51, Christine Lasset78, Min Hyuk Lee79, Jong Won Lee80, Jihyoun Lee79, Jenny Lester71, Fabienne Lesueur81-84, Jennifer T. Loud85, Jan Lubinski57, Noura Mebirouk81-84, Hanne E.J. Meijers-Heijboer86, Alfons Meindl87, Austin Miller88, Marco Montagna17, Thea M. Mooij89, Patrick J. Morrison90, Emmanuelle Mouret-Fourme25, Katherine L. Nathanson30, Susan L. Neuhausen91, Heli Nevanlinna92, Dieter Niederacher93, Finn C. Nielsen94, Robert L. Nussbaum95, Kenneth Offit96, 97, Edith Olah98, Kai-ren Ong99, Laura Ottini100, Sue K. Park101-103, Paolo Peterlongo104, Georg Pfeiler105, Catherine M. Phelan106, Bruce Poppe38, Nisha Pradhan96, Paolo Radice107, Susan J. Ramus108, 109, Johanna Rantala110, Mark Robson97, Gustavo C. Rodriguez111, Rita K. Schmutzler58-

60, Christina Selkirk65, Payal D. Shah30, Jacques Simard112, Christian F. Singer113, Johanna Sokolowska114, Dominique Stoppa-Lyonnet25, 115, 116, Christian Sutter117, Yen Yen Tan28, Manuel R. Teixeira118, 119, Soo H. Teo120, 121, Mary Beth Terry122, Mads Thomassen123, Marc Tischkowitz124, 125, Amanda E. Toland126, Katherine M. Tucker127, 128, Nadine Tung129, Christi J. van Asperen130, Klaartje van Engelen11, Elizabeth J. van Rensburg46, Shan Wang-Gohrke131, Barbara Wappenschmidt58-60, Jeffrey N. Weitzel132, Drakoulis Yannoukakos133, GEMO Study Collaborators 115, 116, HEBON 134, EMBRACE 4, Mark H. Greene85, Matti A. Rookus89, Douglas F. Easton4, 135, Georgia Chenevix-Trench136, Antonis C. Antoniou4, David E. Goldgar137, Olufunmilayo I. Olopade2, Timothy R. Rebbeck138, 139, Dezheng Huo2, 140,

1 Department of Medicine, University of Chicago, Chicago, IL, USA.2 Center for Clinical Cancer Genetics, The University of Chicago, Chicago, IL, USA.3 Division of Gastroenterology, Department of Hepatology and Nutrition, Children's Hospital of Philadelphia, Philadelphia, PA, USA.

2

4 Centre for Cancer Genetic Epidemiology, Department of Public Health and Primary Care, University of Cambridge, Cambridge, UK.5 Department of Genetics, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands.6 Departments of Pediatrics and Medicine, Columbia University, New York, NY, USA.7 Genomic Medicine, Manchester Academic Health Sciences Centre, Division of Evolution and Genomic Sciences, University of Manchester, Central Manchester University Hospitals NHS Foundation Trust, Manchester, UK.8 Institute for Medical Informatics, Statistics and Epidemiology, University of Leipzig, Leipzig, Germany.9 LIFE - Leipzig Research Centre for Civilization Diseases, University of Leipzig, Leipzig, Germany.10 Department of Gynecology and Obstetrics, Technical University of Dresden, Dresden, Germany.11 Department of Clinical Genetics, Academic Medical Center, Amsterdam, The Netherlands.12 Family Cancer Clinic, The Netherlands Cancer Institute - Antoni van Leeuwenhoek hospital, Amsterdam, The Netherlands.13 Yorkshire Regional Genetics Service, Chapel Allerton Hospital, Leeds, UK.14 Department of Pathology, National Institute of Oncology, Budapest, Hungary.15 School of Medicine, University of Iceland, Reykjavik, Iceland.16 Department of Clinical Genetics, Helsinki University Hospital, University of Helsinki, Helsinki, Finland.17 Immunology and Molecular Oncology Unit, Veneto Institute of Oncology  IOV - IRCCS, Padua, Italy.18 Fred A. Litwin Center for Cancer Genetics, Lunenfeld-Tanenbaum Research Institute of Mount Sinai Hospital, Toronto, ON, Canada.19 Department of Molecular Genetics, University of Toronto, Toronto, ON, Canada.20 Department of Breast Medical Oncology, University of Texas MD Anderson Cancer Center, Houston, TX, USA.21 Department of Medical Genetics, University Medical Center, Utrecht, The Netherlands.22 Unit of Medical Genetics, Department of Medical Oncology and Hematology, Fondazione IRCCS (Istituto Di Ricovero e Cura a Carattere Scientifico) Istituto Nazionale dei Tumori (INT), Milan, Italy.23 Oncogénétique, Institut Bergonié, Bordeaux, France.24 Leicestershire Clinical Genetics Service, University Hospitals of Leicester NHS Trust, Leicester, UK.25 Service de Génétique, Institut Curie, Paris, France.26 Human Cancer Genetics Program, Spanish National Cancer Research Centre, Madrid, Spain.27 Centro de Investigación en Red de Enfermedades Raras (CIBERER), Valencia, Spain.28 Dept of OB/GYN, Medical University of Vienna, Vienna, Austria.29 Department of Oncology, Lund University and Skåne University Hospital, Lund, Sweden.30 Department of Medicine, Abramson Cancer Center, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA.

3

31 Genetic Counseling Unit, Hereditary Cancer Program, IDIBGI (Institut d'Investigació Biomèdica de Girona), Catalan Institute of Oncology, CIBERONC, Girona, Spain.32 Department of Medicine, Huntsman Cancer Institute, Salt Lake City, UT, USA.33 Molecular Oncology Laboratory, Hospital Clinico San Carlos, IdISSC, CIBERONC. Martin Lagos s/n, Madrid, Spain.34 Section of Genetic Oncology, Dept. of Laboratory Medicine, University and University Hospital of Pisa, Pisa, Italy.35 Peter MacCallum Cancer Center, Melbourne, Victoria, Australia.36 Sir Peter MacCallum Department of Oncology, The University of Melbourne, Melbourne, Victoria, Australia.37 Unité de recherche en santé des populations, Centre des maladies du sein Deschênes-Fabia, Hôpital du Saint-Sacrement, Québec, QC, Canada.38 Centre for Medical Genetics, Ghent University, Gent, Belgium.39 Department of Clinical Genetics, Erasmus University Medical Center, Rotterdam, The Netherlands.40 Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, USA.41 Unité d'Oncogénétique, CHU Arnaud de Villeneuve, Montpellier, France.42 Department of Clinical Genetics, Fox Chase Cancer Center, Philadelphia, PA, USA.43 Department of Clinical Genetics, South Glasgow University Hospitals, Glasgow, UK.44 Oncogenetics Group, Clinical and Molecular Genetics Area, Vall d'Hebron Institute of Oncology (VHIO), University Hospital, Vall d'Hebron, Barcelona, Spain.45 Clinical Genetics Department, St Michael's Hospital, Bristol, UK.46 Department of Genetics, University of Pretoria, Arcadia, South Africa.47 Oncogenetics Team, The Institute of Cancer Research and Royal Marsden NHS Foundation Trust, London, UK.48 Molecular Diagnostic Unit, Hereditary Cancer Program, ICO-IDIBELL (Catalan Institute of Oncology, Bellvitge Biomedical Research Institute), CIBERONC, Barcelona, Spain.49 Department of Cancer Epidemiology and Genetics, Masaryk Memorial Cancer Institute, Brno, Czech Republic.50 Ohio State University Columbus Cancer Council, Columbus, OH, USA.51 The Susanne Levy Gertner Oncogenetics Unit, Chaim Sheba Medical Center, Ramat Gan, Israel.52 Sackler Faculty of Medicine, Tel Aviv University, Ramat Aviv, Israel.53 Schools of Medicine and Public Health, Division of Cancer Prevention & Control Research, Jonsson Comprehensive Cancer Centre, UCLA, Los Angeles, CA, USA.54 Cancer Risk and Prevention Clinic, Dana-Farber Cancer Institute, Boston, MA, USA.55 Department of Pathology and Laboratory Medicine, Kansas University Medical Center, Kansas City, KS, USA.56 Department of Clinical Genetics and GROW, School for Oncology and Developmental Biology, Maastricht University Medical Center, Maastricht, The Netherlands.57 Department of Genetics and Pathology, Pomeranian Medical University, Szczecin, Poland.58 Center for Hereditary Breast and Ovarian Cancer, University Hospital of Cologne, Cologne, Germany.

4

59 Center for Integrated Oncology (CIO), University Hospital of Cologne, Cologne, Germany.60 Center for Molecular Medicine Cologne (CMMC), University of Cologne, Cologne, Germany.61 Molecular Genetics of Breast Cancer, German Cancer Research Center (DKFZ), Heidelberg, Germany.62 Institute of Genetic Medicine, Centre for Life, Newcastle Upon Tyne Hospitals NHS Trust, Newcastle upon Tyne, UK.63 City of Hope Clinical Cancer Genetics Community Research Network, Duarte, CA, USA.64 Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Victoria, Australia.65 Center for Medical Genetics, NorthShore University HealthSystem, Evanston, IL, USA.66 The University of Chicago Pritzker School of Medicine, Chicago, IL, USA.67 N.N. Petrov Institute of Oncology, St. Petersburg, Russia.68 Lombardi Comprehensive Cancer Center, Georgetown University, Washington, DC, USA.69 Clinical Genetics, Guy's and St Thomas' NHS Foundation Trust, London, UK.70 Department of Surgery, Seoul National University Bundang Hospital, Seongnam, Korea.71 Women's Cancer Program at the Samuel Oschin Comprehensive Cancer Institute, Cedars-Sinai Medical Center, Los Angeles, CA, USA.72 Department of Human Genetics, Radboud University Medical Center, Nijmegen, The Netherlands.73 Department of Surgery, Daerim Saint Mary's Hospital, Seoul, Korea.74 Department of Surgery, Soonchunhyang University Hospital Bucheon Hospital, Bucheon, Korea.75 Hong Kong Hereditary Breast Cancer Family Registry, Happy Valley, Hong Kong.76 Department of Surgery, The University of Hong Kong, Pok Fu Lam, Hong Kong.77 Department of Surgery, Hong Kong Sanatorium and Hospital, Happy Valley, Hong Kong.78 Unité de Prévention et d’Epidémiologie Génétique, Centre Léon Bérard, Lyon, France.79 Department of Surgery, Soonchunhyang University College of Medicine and Soonchunhyang University Hospital, Seoul, Korea.80 Department of Surgery, Ulsan University College of Medicine and Asan Medical Center, Seoul, Korea.81 Genetic Epidemiology of Cancer team, Institut Curie, Paris, France.82 U900, INSERM, Paris, France.83 PSL University, Paris, France.84 Mines ParisTech, Fontainebleau, France.85 Clinical Genetics Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA.86 Department of Clinical Genetics, VU University Medical Center, Amsterdam, The Netherlands.87 Division of Gynaecology and Obstetrics, Technische Universität München, Munich, Germany.88 NRG Oncology, Statistics and Data Management Center, Roswell Park Cancer Institute, Buffalo, NY, USA.89 Department of Epidemiology, The Netherlands Cancer Institute, Amsterdam, The Netherlands.90 Centre for Cancer Research and Cell Biology, Queens University Belfast, Belfast, UK.

5

91 Department of Population Sciences, Beckman Research Institute of City of Hope, Duarte, CA, USA.92 Department of Obstetrics and Gynecology, Helsinki University Hospital, University of Helsinki, Helsinki, Finland.93 Department of Gynecology and Obstetrics, University Hospital Düsseldorf, Heinrich-Heine University Düsseldorf, Düsseldorf, Germany.94 Center for Genomic Medicine, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark.95 Cancer Genetics and Prevention Program, University of California San Francisco, San Francisco, CA, USA.96 Clinical Genetics Research Lab, Department of Cancer Biology and Genetics, Memorial Sloan-Kettering Cancer Center, New York, NY, USA.97 Clinical Genetics Service, Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, NY, USA.98 Department of Molecular Genetics, National Institute of Oncology, Budapest, Hungary.99 West Midlands Regional Genetics Service, Birmingham Women’s Hospital Healthcare NHS Trust, Birmingham, UK.100 Department of Molecular Medicine, University La Sapienza, Rome, Italy.101 Department of Preventive Medicine, Seoul National University College of Medicine, Seoul, Korea.102 Department of Biomedical Sciences, Seoul National University Graduate School, Seoul, Korea.103 Cancer Research Institute, Seoul National University, Seoul, Korea.104 IFOM, The FIRC (Italian Foundation for Cancer Research) Institute of Molecular Oncology, Milan, Italy.105 Department of Urology, Medical University of Vienna, Vienna, Austria.106 Department of Cancer Epidemiology, Moffitt Cancer Center, Tampa, FL, USA.107 Unit of Molecular Bases of Genetic Risk and Genetic Testing, Department of Preventive and Predictive Medicine, Fondazione IRCCS (Istituto Di Ricovero e Cura a Carattere Scientifico) Istituto Nazionale dei Tumori (INT), Milan, Italy.108 School of Women's and Children's Health, University of New South Wales Sydney, New South Wales, Australia.109 The Kinghorn Cancer Centre, Garvan Institute of Medical Research, Sydney, New South Wales, Australia.110 Clinical Genetics, Karolinska Institutet, Stockholm, Sweden.111 Division of Gynecologic Oncology, NorthShore University HealthSystem, University of Chicago, Evanston, IL, USA.112 Genomics Center, Centre Hospitalier Universitaire de Québec Research Center, Laval University, Québec City, QC, Canada.113 Dept of OB/GYN and Comprehensive Cancer Center, Medical University of Vienna, Vienna, Austria.114 Laboratoire de génétique médicale, Nancy Université, Centre Hospitalier Régional et Universitaire, Vandoeuvre-les-Nancy, France.

6

115 Department of Tumour Biology, Institut Curie, INSERM U830, Paris, France.116 Université Paris Descartes, INSERM UMR-S1147, Paris, France.117 Institute of Human Genetics, University Hospital Heidelberg, Heidelberg, Germany.118 Department of Genetics, Portuguese Oncology Institute, Porto, Portugal.119 Biomedical Sciences Institute (ICBAS), University of Porto, Porto, Portugal.120 Cancer Research Malaysia, Subang Jaya, Selangor, Malaysia.121 Breast Cancer Research Unit, Cancer Research Institute, University Malaya Medical Centre, Kuala Lumpur, Malaysia.122 Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY, USA.123 Department of Clinical Genetics, Odense University Hospital, Odence C, Denmark.124 Program in Cancer Genetics, Departments of Human Genetics and Oncology, McGill University, Montréal, QC, Canada.125 Department of Medical Genetics, Addenbrooke's Hosptital, Cambridge, UK.126 Department of Cancer Biology and Genetics, Comprehensive Cancer Center, The Ohio State University, Columbus, OH, USA.127 School of Medicine, University of NSW, Sydney, New South Wales, Australia.128 Hereditary Cancer Centre, Prince of Wales Hospital, Sydney, NSW, Australia.129 Department of Medical Oncology, Beth Israel Deaconess Medical Center, Boston, MA, USA.130 Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands.131 Department of Gynaecology and Obstetrics, University of Ulm, Ulm, Germany.132 Clinical Cancer Genetics, City of Hope, Duarte, CA, USA.133 Molecular Diagnostics Laboratory, INRASTES, National Centre for Scientific Research 'Demokritos', Athens, Greece.134 The Hereditary Breast and Ovarian Cancer Research Group Netherlands (HEBON), Coordinating center: The Netherlands Cancer Institute, Amsterdam, The Netherlands.135 Centre for Cancer Genetic Epidemiology, Department of Oncology, University of Cambridge, Cambridge, UK.136 Department of Genetics and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia.137 Department of Dermatology, Huntsman Cancer Institute, University of Utah School of Medicine, Salt Lake City, UT, USA.138 Harvard T.H. Chan School of Public Health, Boston, MA, USA.139 Dana-Farber Cancer Institute, Boston, MA, USA.140 Department of Public Health Sciences, University of Chicago, Chicago, IL, USA.

Corresponding author: Dr. Dezheng Huo, 5841 S. Maryland Ave., MC 2000, Chicago, Illinois 60637. Phone: (773) 834-0843; E-mail: dhuo@health.bsd.uchicago.edu

Keywords: BRCA1/2, Mendelian Randomization, Breast Cancer, Body Height, Body Mass

Index

Word limits: Abstract 250 (Limits 250); Text 3154 (Limit 3000)

7

Abstract

Background BRCA1/2 mutations confer a high life-time risk of developing breast cancer, but

this risk may be modified by other factors. Whether height or body mass index (BMI) modify

breast cancer risk in BRCA1/2 mutation carriers remain unclear.

Methods We used Mendelian randomization approaches to evaluate the association of height

and BMI on breast cancer risk, using data from the Consortium of Investigators for the Modifiers

of BRCA1/2 with 14,676 BRCA1 and 7,912 BRCA2 mutation carriers, including 11,451 cases of

breast cancer. We created a height genetic score (height-GS) using 586 height-associated

variants and a BMI genetic score (BMI-GS) using 93 BMI-associated variants. We examined

both observed and genetically determined height and BMI in relation to breast cancer risk using

weighted Cox models.

Results Observed height was positively associated with breast cancer risk (hazard ratio [HR]:

1.09 per 10 cm increase, 95% confidence interval [CI]: 1.02–1.17; P=0.016). Height-GS was

positively associated with breast cancer, though this was not statistically significant (HR: 1.04

per 10 cm increase in genetically predicted height, 95% CI: 0.93–1.17; P=0.47). Observed BMI

was inversely associated with breast cancer risk (HR: 0.94 per 5 kg/m2 increase, 95% CI: 0.90–

0.98; P=0.007). BMI-GS was also inversely associated with breast cancer risk (HR: 0.87 per 5

kg/m2 increase in genetically predicted BMI, 95% CI: 0.76–0.98; P=0.023).

Conclusion Body height and weight were associated with breast cancer risk in BRCA1/2 carriers

and incorporating anthropometric measures into risk assessment could improve the cancer risk

management in this population.

8

Introduction

Breast cancer is the most common cancer in women and one of the leading causes of

cancer deaths globally [1]. Inheritance of a BRCA1 or BRCA2 mutation is associated with an

increased lifetime risk of developing breast cancer [2, 3]. However, penetrance of BRCA1/2

mutations is likely to be modified by lifestyle, reproductive factors, and other genetic variants [4-

8]. Multiple genes have been found to modify the association between BRCA1/2 and breast

cancer risk [9-11]. Accurate breast cancer risk prediction in BRCA1/2 mutation carriers is crucial

in preventing morbidity and mortality, while optimizing the timeline of prophylactic procedures.

The relationship between anthropometric characteristics such as height or body mass

index (BMI) and breast cancer risk have been extensively studied [12, 13]. Adult height was

found to have a positive association with breast cancer risk [14]. Higher BMI is positively

associated with postmenopausal breast cancer and inversely associated with premenopausal

breast cancer [15]. However, the associations of height and BMI with breast cancer risk in

BRCA1/2 mutation carriers remain unclear. Retrospective studies are subject to potential biases,

while prospective studies are often underpowered.

Notably, both height and BMI have a strong genetic basis. Genome-wide association

studies (GWAS) [16-18] have identified numerous variants that are associated with either trait.

In aggregate, these variants explain a significant fraction of the variation in each trait.

Mendelian randomization (MR) is a method that assesses the association between an

exposure and disease by using genetic markers associated with the exposure as instrumental

variables (IVs). Since genes are inherited randomly, MR can be used to minimize the effects of

recall bias, reverse causation, measurement error and residual confounding [19]. The

assumptions underlying MR include: genetic variants are associated with the exposure of

interest, genetic variants only affect the outcome through the exposure, and the genetic variants

are weakly or not associated with confounders in the exposure-outcome pathway [20, 21]. A

causal relationship between exposure and disease could be concluded if these assumptions are

held. In this study, we employed a MR approach to evaluate the association between height and

BMI and breast cancer, using data from the Consortium of Investigators of Modifiers of

BRCA1/2 (CIMBA), including 22,588 women, with 14,676 BRCA1 and 7,912 BRCA2 mutation

carriers.

9

Methods

Information about CIMBA and the genotyping protocols can be found in the

Supplementary Methods and previous publications in detail [9-11].

SNP Selection

SNPs associated with height and BMI were identified from the Genetic Investigation of

Anthropometric Traits publications [16, 17]. SNPs associated with height and BMI at genome-

wide significance (P<5×10−8) were eligible. We excluded SNPs with an imputation quality <0.5.

For height, we included 586 SNPs (85 genotyped) (Supplementary Table 1). For BMI, we

included 93 SNPs (12 genotyped) (Supplementary Table 2).

Statistical Analysis

We calculated weighted genetic scores (GS) for height and BMI using methods described

previously, based on a polygenic additive model (i.e. ignoring interactions between variants) [14,

22]. We calculated each GS using the formulas: GSheight=∑i=1

586

β i SNPi andGSBMI=∑i=1

93

β i SNPi,

where βi is the per-allele effect of the ith SNP for height and BMI reported in previous studies

[16, 17] and SNPi is the dosage of the effect allele (0, 1, or 2) of the ith SNP. We re-scaled the

GSs such that the score provided a genetically predicted height and BMI, by performing linear

regressions of observed height and BMI on the corresponding GS in non-cases. For height, we

obtained from the regression equation β0 (intercept=165.648) and β1 (slope=5.119). The

corresponding values for BMI were β0 (22.058) and β1 (6.408). We used these values to calculate

the scaled height- and BMI-GS using the equation: Scaled-GS=β0+β1GS. We also estimated the

percent variation explained by each GS and assessed the correlation between each GS and

traditional risk factors for breast cancer.

Next, we modeled height- and BMI-GS with breast cancer risk using weighted Cox

models. The primary outcome was diagnosis of breast cancer. Observations were censored at

diagnosis of ovarian cancer, prophylactic mastectomy/salpingo-oophorectomy, death, or the end

of follow-up, whichever came first. Time to event was computed from birth to age at breast

cancer diagnosis or censoring. Mutation carriers were not randomly selected and usually those

10

with breast cancer had a higher probability to do genetic testing. To account for this non-random

sampling, we applied a weighted cohort approach as described previously [23]. Weights were

assigned based on previously observed incidence rates of breast cancer for BRCA1/2 carriers

[24]. To account for interdependence between related carriers from the same family, we used a

robust sandwich variance-estimation approach. Stratified analyses were performed by BRCA1/2

or menopausal status. Menopausal status was modeled as a time-varying covariate: the variable

was coded as premenopausal from birth until age at censoring and was switched to

postmenopausal at the age of natural menopause or bilateral salpingo-oophorectomy. If age at

natural menopause or bilateral salpingo-oophorectomy was missing, we imputed the age as 46,

since this was the mean menopausal age in this population. The analyses were also adjusted for

the first eight principal components (as a proxy for population structure and ethnicity) and birth

cohort and stratified by country of enrollment.

We also examined the association between height and BMI with breast cancer by

modeling each individual height and BMI variant separately. We assessed the direct association

between each SNP and height and BMI (βXG) and its association with risk of breast cancer (βYG).

βXG for each SNP was extracted from prior GWAS publications and represents the per-allele

effect on either height or BMI. βYG was calculated using multivariable-adjusted weighted Cox

model for each SNP using data from CIMBA, i.e. breast cancer ~ βYGX (where X = 0, 1, 2 for the

allele corresponding to increased height or BMI), principal components, birth cohort, mutated

gene, and country of enrollment. We statistically combined these two effect estimates to measure

the association between height and BMI and breast cancer risk (βYX) [25, 26]. The causal effect

(βYX) was calculated using the Wald estimator:βYX=βYG

βXG. The standard error for this estimate was

estimated using the method proposed by Burgess and colleagues: SEYX=√( SEYG

β XG)

2

[26]. βYX can

be interpreted as log hazard ratio (HR) for breast cancer per 1 unit increase in genetically

determined height and BMI. We then combined the effects of individual height and BMI-

associated variants using an inverse-variance fixed-effects meta-analysis. We also used Egger’s

test to assess for possible pleiotropic effects of the variants (i.e. the effects are not mediated via

the exposure), one of the assumptions for MR [27].

11

In the subset of participants with reported height or BMI (34%), we performed a formal

instrumental variable analysis to estimate the unbiased effect of height and BMI on breast cancer

risk using two-stage residual inclusion regression [28]. In stage one, we conducted a linear

regression of observed height or BMI ~ corresponding GS, principal components, birth cohort,

country, mutated genes, and residuals. In stage two, we used a Cox model to fit cancer risk

against height and BMI, birth cohort, country, mutation status, and residuals from stage one. We

performed 10,000 boot-straps to obtain correct variance estimates.

Lastly, we examined the association between reported height and BMI and breast cancer

risk in participants with reported height and BMI using weighted Cox models, adjusted for

traditional breast cancer risk factors including birth cohort, menopausal status, age at menarche

(continuous) and parity (continuous). BMI was reported at date of questionnaire (at baseline),

usually close to the date of genetic testing and recalled for young adulthood (at age 18). The

BMI-GS mentioned above was rescaled to BMI reported at baseline because previous GWAS

were not based on BMI in young adulthood.

The proportional hazards assumption was tested by adding an interaction term of age and

either height-GS or BMI-GS. In models with menopausal status as time-varying variable, test for

heterogeneity by menopausal status was also a test for proportional hazard assumption. Analyses

were performed using SAS 9.4 (SAS Institute, Cary, NC) and Stata 14.0 (StataCorp, College

Station, TX). A two-sided P-value<0.05 was considered statistically significant unless stated

otherwise.

Results

Baseline Characteristics

Table 1 presents the baseline characteristics of the 22,588 participants (14,676 BRCA1

and 7,912 BRCA2 mutation carriers) with genotype information. The mean age of individuals at

the time of breast cancer diagnosis (cases) was similar to the age of individuals who did not

develop breast cancer at the time of censoring (controls). However, the birth year of cases tended

to be earlier than controls. Height was available in 7,657 participants (4,502 BRCA1 carriers and

3,155 BRCA2 carriers) and BMI at date of questionnaire was available in 7,516 participants

(4,401 BRCA1 carriers and 3,115 BRCA2 carriers).

12

Height Analysis

Observed height was positively associated with breast cancer risk after multivariable

adjustment (HR per 10 cm: 1.09, 95% CI: 1.02–1.17, P=0.016) (Table 2). In stratified analysis,

we found that height was a stronger predictor of risk in BRCA2 carriers (HR=1.17, 95% CI:

1.04–1.31) than in BRCA1 carriers (HR=1.06, 95% CI: 0.97–1.16), but the interaction was not

statistically significant. The country-specific estimates showed low levels of heterogeneity

(Supplementary Figure 1a).

We found that height-GS was strongly associated with observed height by case/control

and mutation status (all P<10-93) (Table 3). The height-GS explained 13.4% of the variation in

height (Supplementary Figure 2a), which is consistent with estimates from previous GWAS of

height [17]. As shown in Supplementary Figure 3a, there was a strong correlation (r=0.44)

between the estimated effect size for individual variants in our study and those reported in

previous GWAS. Height-GS was positively associated with weight, baseline age, and age at

menarche but the associations were weak.

Height-GS was positively associated with breast cancer risk with an effect weaker than

that for observed height, though it was not statistically significant (HR: 1.04 per 10 cm increase

in genetically predicted height, 95% CI: 0.93–1.17; P=0.47) (Table 4). Effect was not different

when stratified for menopausal or mutation status.

When combining the breast cancer risk estimates for individual height-variant using

inverse-variance meta-analysis, the result was similar (HR: 1.05, 95% CI: 0.93–1.19; P=0.42)

(Table 4). There was low heterogeneity among SNPs (I2=17.0%). Egger’s test for small-study

effects was not significant (P=0.61, Supplementary Figure 4a), so we failed to reject the

assumption of no pleiotropic effects for MR analysis. The two-stage residual inclusion analysis

found a similar risk estimate to that for observed height (HR: 1.09, 95% CI: 0.93–1.27; P=0.27)

(Table 4).

BMI Analysis

For reported BMI at date of questionnaire, we found an inverse association with breast

cancer risk after multivariable-adjustment (HR per 5 kg/m2 increase: 0.94, 95% CI: 0.90–0.98;

P=0.007) (Table 5). Association was stronger in BRCA2 vs. BRCA1 carriers (HR: 0.90 vs. 0.96)

and for premenopausal vs. postmenopausal breast cancer (HR: 0.92 vs. 0.97), but there was no

13

statistically significant interaction (Pinteraction>0.05 for each comparison). We found a stronger

inverse association of BMI in young adulthood with breast cancer risk (HR: 0.83, 95% CI: 0.76–

0.90; P=2.1x10-5). The country-specific estimates showed moderate levels of heterogeneity

(Supplementary Figure 1b).

BMI-GS was strongly associated with reported BMI at date of questionnaire among

controls and cases (each P<10-14) (Table 6). BMI-GS accounted for 2.6% of the variation in BMI

at date of questionnaire (Supplementary Figure 2b). We found a strong correlation between the

effect on BMI by individual variants in prior reported GWAS and in CIMBA (r=0.52,

Supplementary Figure 3b). Similarly, the BMI-GS was associated with reported BMI in young

adulthood, with stronger effects among controls (P<10-15, r2=2.3%). The BMI-GS was positively

associated with height and inversely associated with age at menarche.

In the analysis of BMI-GS and breast cancer risk, each 5 kg/m2 increment in genetically

predicted BMI was associated with a 13% reduction in breast cancer risk (HR: 0.87, 95% CI:

0.76–0.98; P=0.023) (Table 7). The association was slightly stronger among BRCA2 mutation

carriers and for premenopausal breast cancer, though there was no significant interaction

(Pinteraction>0.05 for each).

When we statistically combined the effect of individual BMI-variants on breast cancer

risk, we found a similar association with risk of breast cancer (HR: 0.87, 95% CI: 0.76–0.98,

P=0.027) (Table 7). There was low heterogeneity in the SNPs (I2=3.5%). Egger’s test was not

significant (P=0.44, Supplementary Figure 4b), suggesting that pleiotropic effects may not

exist. The two-stage residual inclusion method yielded similar results (HR: 0.86, 95% CI: 0.70–

1.07; P=0.18) (Table 7).

Individual Height- and BMI-associated Variants

Of the 586 height-related variants, 50 were found to be associated with breast cancer risk

at P<0.05 (Table 8). Of the 93 BMI-related variants, 7 were associated with breast cancer risk.

One SNP (rs10744956) was statistically significant after Bonferroni adjustment.

Discussion

Using data from a large international study of women with BRCA1/2 mutations and

analyzed by several Mendelian randomization methods, we found that both observed and

14

genetically-predicted BMI were associated with a reduced risk of breast cancer whereas observed

and genetically-predicted height was associated with an increased risk of breast cancer.

We found that each 10-cm increment in observed height was associated with a 9%

increase in breast cancer risk, whereas a 10-cm increment in genetically predicted height was

associated with a 4-8% increase in risk in BRCA1/2 mutation carriers. Our findings are broadly

consistent with previous studies in the general population [14, 20]. A recent meta-analysis of

prospective studies of height reported a relative risk (RR) of 1.17 per 10 cm increase, and the

MR analysis using 168 height-associated variants found an odds ratio (OR) of 1.22 per 10 cm

increase in genetically predicted height [14]. A subsequent MR analysis with 423 height

associated-variants reported a similar result (OR=1.19) [20]. One study of height in 719

BRCA1/2 mutation carriers showed a non-significant positive relationship with premenopausal

breast cancer and a significant positive relationship with postmenopausal breast cancer [29].

Thus, height is likely a predictor for breast cancer risk in BRCA1/2 mutation carriers and the

general population.

Several studies have examined the relationship between BMI and breast cancer risk in

BRCA1/2 carriers [5, 29, 30] with inconsistent findings, possibly because of limited sample size.

In the general population, every 5 kg/m2 increase in BMI was positively associated with

postmenopausal breast cancer (RR=1.12) and inversely associated with premenopausal breast

cancer (RR=0.92) [15]. We found that for BRCA1/2 carriers, observed BMI at date of

questionnaire was inversely associated with premenopausal breast cancer (HR=0.92) but was not

significantly associated with postmenopausal breast cancer (HR=0.97). Our MR analysis found

that a 5 kg/m2 increase in genetically predicted BMI was associated with a 16% reduction in

premenopausal breast cancer and a 11% reduction in postmenopausal breast cancer. Similarly, a

MR analysis in the general population found that each 5 kg/m2 increase in genetically predicted

BMI had an OR of 0.65, with consistent effects across menopausal status [22]. Altogether, there

is strong evidence for the protective effect of higher BMI on premenopausal breast cancer in

both the general population and BRCA1/2 mutation carriers. Unlike with MR, the association of

breast cancer risk with observed BMI is potentially subject to recall bias or reverse causation.

Conversely, BMI-GS may only capture early life body weight and cannot predict weight changes

later in life, which is influenced more by lifestyle factors. Interestingly, the association between

BMI at age 18 and premenopausal breast cancer (HR=0.83) was quite similar to that for BMI-GS

15

and premenopausal breast cancer (HR=0.84), supporting the notion that early life BMI or

adiposity may play a key role in breast carcinogenesis. The seemingly inconsistent findings on

the relationship between BMI and postmenopausal breast cancer might reflect differences in

study populations and methodology. Since the mean age at breast cancer diagnosis was 41.1 for

BRCA1 carriers and 44.2 for BRCA2 carriers, few of the cases occurred after menopause. In turn,

it probably takes time for the protective effect of high BMI to disappear after menopause.

There are several potential mechanisms for the associations between height and BMI and

breast cancer. For height, early life exposures including nutritional and hormonal status could

exert an impact on obtained height and explain the association between obtained height and

breast cancer risk [31, 32]. The insulin-like growth factor (IGF) signaling pathway has been

implicated in the pathogenesis of multiple malignancies, including breast cancer, with possibly

stronger effects on premenopausal breast cancer [33, 34]. Recent investigations have also

implicated the LIN28B-let-7 microRNA pathway, which are regulators of adult height,

mammalian body size, and carcinogenesis [35-37]. Furthermore, potential mechanisms that

could account for the association between BMI and reduced risk of breast cancer include

circulating IGF-1 [15], greater likelihood of having anovulatory cycles, and lower circulating

levels of estradiol/progesterone [38].

We were able to replicate associations for several SNPs reported in previous studies in

the general population. For BMI, Guo et al [22] found that rs7903146 near the TCF7L2 gene

(OR=0.96) and rs1558902 (OR=0.93) near the FTO gene were associated with breast cancer risk.

These estimates were consistent with our findings (rs7903146: HR=0.96; rs1558902: HR=0.97).

Interestingly, rs7903146 is in weak linkage disequilibrium (LD) (r2=0.45) with rs7904519 near

TCF7L2, which was reported in a previous GWAS of breast cancer [39]. Moreover, rs1558902

was in strong LD (r2 =0.92) with rs17817449 near FTO, which was reported in previous GWAS

[39, 40]. The finding of a significant association with a SNP near FTO further potentiates its

involvement in breast cancer [41-43]. Functional analyses are warranted to better understand the

mechanisms in which FTO and TCF7L2 affect breast cancer risk.

The strengths of our study include a large sample size, inclusion of numerous height and

BMI-variants, MR approach which is less likely to be confounded by other risk factors, and

consistent findings between observed and genetically predicted phenotypes. Our study has

several limitations. Observed height and BMI for breast cancer cases were typically measured

16

approximately 5-6 years after initial diagnosis. While height is unlikely to be affected by breast

cancer diagnosis, breast cancer-related weight loss may affect the relationship between observed

BMI and breast cancer risk. The height-GS explained 13.4% of height variation, similar to that

reported in previous GWAS (15.9%) [17]. The BMI-GS accounted for 2.6% of BMI variation,

similar to the 2.7% reported in previous GWAS [16, 17]. Although both GSs had sufficient

strength to be valid IVs (F-statistic>>10), they are not very strong IVs, leading to wide

confidence intervals in the MR analysis. Although the GSs were correlated with some breast

cancer risk factors, these associations were much weaker when compared to height or BMI,

suggesting minimal residual confounding and upholding the assumptions of MR. Another

limitation is that our study only included women of European ancestry, which limits

generalizability to women of other racial/ethnic groups.

In summary, our study supports that higher BMI is associated with lower risk of breast

cancer whereas greater height may be associated with increased risk of breast cancer for women

with BRCA1/2 mutations. These findings may have implications for clinical risk stratification to

help carriers and their physicians to decide age-appropriate risk-tailored interventions, including

increased surveillance and prophylactic surgeries. Future studies could elucidate the biological

mechanisms underlying height and BMI and breast cancer risk.

Acknowledgements

CIMBA: The CIMBA data management and data analysis were supported by Cancer Research – UK grants C12292/A20861, C12292/A11174. GCT and ABS are NHMRC Research Fellows. iCOGS: the European Community's Seventh Framework Programme under grant agreement n° 223175 (HEALTH-F2-2009-223175) (COGS), Cancer Research UK (C1287/A10118, C1287/A 10710, C12292/A11174, C1281/A12014, C5047/A8384, C5047/A15007, C5047/A10692, C8197/A16565), the National Institutes of Health (CA128978) and Post-Cancer GWAS initiative (1U19 CA148537, 1U19 CA148065 and 1U19 CA148112 - the GAME-ON initiative), the Department of Defence (W81XWH-10-1-0341), the Canadian Institutes of Health Research (CIHR) for the CIHR Team in Familial Risks of Breast Cancer (CRN-87521), and the Ministry of Economic Development, Innovation and Export Trade (PSR-SIIRI-701), Komen Foundation for the Cure, the Breast Cancer Research Foundation, and the Ovarian Cancer Research Fund. The PERSPECTIVE project was supported by the Government of Canada through Genome Canada and the Canadian Institutes of Health Research, the Ministry of Economy, Science and Innovation through Genome Québec, and The Quebec Breast Cancer Foundation.

BCFR: UM1 CA164920 from the National Cancer Institute. The content of this manuscript does not necessarily reflect the views or policies of the National Cancer Institute or any of the

17

collaborating centers in the Breast Cancer Family Registry (BCFR), nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government or the BCFR. BFBOCC: Lithuania (BFBOCC-LT): Research Council of Lithuania grant SEN-18/2015. BIDMC: Breast Cancer Research Foundation. BMBSA: Cancer Association of South Africa (PI Elizabeth J. van Rensburg). CNIO: Spanish Ministry of Health PI16/00440 supported by FEDER funds, the Spanish Ministry of Economy and Competitiveness (MINECO) SAF2014-57680-R and the Spanish Research Network on Rare diseases (CIBERER). COH-CCGCRN: Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under grant number R25CA112486, and RC4CA153828 (PI: J. Weitzel) from the National Cancer Institute and the Office of the Director, National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. CONSIT: Associazione Italiana Ricerca sul Cancro (AIRC; IG2014 no.15547) to P. Radice. Italian Association for Cancer Research (AIRC; grant no.16933) to L. Ottini. Associazione Italiana Ricerca sul Cancro (AIRC; IG2015 no.16732) to P. Peterlongo. Jacopo Azzollini is supported by funds from Italian citizens who allocated the 5x1000 share of their tax payment in support of the Fondazione IRCCS Istituto Nazionale Tumori, according to Italian laws (INT-Institutional strategic projects ‘5x1000’). DEMOKRITOS: European Union (European Social Fund – ESF) and Greek national funds through the Operational Program "Education and Lifelong Learning" of the National Strategic Reference Framework (NSRF) - Research Funding Program of the General Secretariat for Research & Technology: SYN11_10_19 NBCA. Investing in knowledge society through the European Social Fund. DFKZ: German Cancer Research Center. EMBRACE: Cancer Research UK Grants C1287/A10118 and C1287/A11990. D. Gareth Evans and Fiona Lalloo are supported by an NIHR grant to the Biomedical Research Centre, Manchester. The Investigators at The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust are supported by an NIHR grant to the Biomedical Research Centre at The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust. Ros Eeles and Elizabeth Bancroft are supported by Cancer Research UK Grant C5047/A8385. Ros Eeles is also supported by NIHR support to the Biomedical Research Centre at The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust. FCCC: The University of Kansas Cancer Center (P30 CA168524) and the Kansas Bioscience Authority Eminent Scholar Program. A.K.G. was funded by R0 1CA140323, R01 CA214545, and by the Chancellors Distinguished Chair in Biomedical Sciences Professorship. FPGMX: FISPI05/2275 and Mutua Madrileña Foundation (FMMA). GC-HBOC: German Cancer Aid (grant no 110837, Rita K. Schmutzler) and the European Regional Development Fund and Free State of Saxony, Germany (LIFE - Leipzig Research Centre for Civilization Diseases, project numbers 713-241202, 713-241202, 14505/2470, 14575/2470). GEMO: Ligue Nationale Contre le Cancer; the Association “Le cancer du sein, parlons-en!” Award, the Canadian Institutes of Health Research for the "CIHR Team in Familial Risks of Breast Cancer" program and the French National Institute of Cancer (INCa). GEORGETOWN: the Non-Therapeutic Subject Registry Shared Resource at Georgetown University (NIH/NCI grant P30-CA051008), the Fisher Center for Hereditary Cancer and Clinical Genomics Research, and Swing For the Cure. G-FAST: Bruce Poppe is a senior clinical investigator of FWO. Mattias Van Heetvelde obtained funding from IWT. HCSC: Spanish Ministry of Health PI15/00059,

18

PI16/01292, and CB-161200301 CIBERONC from ISCIII (Spain), partially supported by European Regional Development FEDER funds. HEBCS: Helsinki University Hospital Research Fund, Academy of Finland (266528), the Finnish Cancer Society and the Sigrid Juselius Foundation. HEBON: the Dutch Cancer Society grants NKI1998-1854, NKI2004-3088, NKI2007-3756, the Netherlands Organization of Scientific Research grant NWO 91109024, the Pink Ribbon grants 110005 and 2014-187.WO76, the BBMRI grant NWO 184.021.007/CP46 and the Transcan grant JTC 2012 Cancer 12-054. HEBON thanks the registration teams of Dutch Cancer Registry (IKNL; S. Siesling, J. Verloop) and the Dutch Pathology database (PALGA; L. Overbeek) for part of the data collection. HRBCP: Hong Kong Sanatorium and Hospital, Dr Ellen Li Charitable Foundation, The Kerry Group Kuok Foundation, National Institute of Health1R 03CA130065, and North California Cancer Center. HUNBOCS: Hungarian Research Grants KTIA-OTKA CK-80745 and OTKA K-112228. ICO: The authors would like to particularly acknowledge the support of the Asociación Española Contra el Cáncer (AECC), the Instituto de Salud Carlos III (organismo adscrito al Ministerio de Economía y Competitividad) and “Fondo Europeo de Desarrollo Regional (FEDER), una manera de hacer Europa” (PI10/01422, PI13/00285, PIE13/00022, PI15/00854, PI16/00563 and CIBERONC) and the Institut Català de la Salut and Autonomous Government of Catalonia (2009SGR290, 2014SGR338 and PERIS Project MedPerCan). Asociación Española Contra el Cáncer, Spanish Health Research Foundation, Carlos III Health Institute, organismo adscrito al Ministerio de Economía y Competitividad, “Fondo Europeo de Desarrollo Regional (FEDER), una manera de hacer Europa”, Catalan Health Institute, and Autonomous Government of Catalonia (ISCIIIRETIC RD06/0020/1051, RD12/0036/008, PI13/00285, PIE13/00022, PI15/00854, PI16/00563 and 2009SGR283). IHCC: PBZ_KBN_122/P05/2004. ILUH: Icelandic Association “Walking for Breast Cancer Research” and by the Landspitali University Hospital Research Fund. INHERIT: Canadian Institutes of Health Research for the “CIHR Team in Familial Risks of Breast Cancer” program – grant # CRN-87521 and the Ministry of Economic Development, Innovation and Export Trade – grant # PSR-SIIRI-701. IOVHBOCS: Ministero della Salute and “5x1000” Istituto Oncologico Veneto grant. IPOBCS: Liga Portuguesa Contra o Cancro. kConFab: The National Breast Cancer Foundation, and previously by the National Health and Medical Research Council (NHMRC), the Queensland Cancer Fund, the Cancer Councils of New South Wales, Victoria, Tasmania and South Australia, and the Cancer Foundation of Western Australia. MAYO: NIH grants CA116167, CA192393 and CA176785, an NCI Specialized Program of Research Excellence (SPORE) in Breast Cancer (CA116201), and a grant from the Breast Cancer Research Foundation. MCGILL: Jewish General Hospital Weekend to End Breast Cancer, Quebec Ministry of Economic Development, Innovation and Export Trade. Marc Tischkowitz is supported by the funded by the European Union Seventh Framework Program (2007Y2013)/European Research Council (Grant No. 310018). MODSQUAD: MH CZ - DRO (MMCI, 00209805), MEYS - NPS I - LO1413 to LF and by the European Regional Development Fund and the State Budget of the Czech Republic (RECAMO, CZ.1.05/2.1.00/03.0101) to LF, and by Charles University in Prague project UNCE204024 (MZ). MSKCC: the Breast Cancer Research Foundation, the Robert and Kate Niehaus Clinical Cancer Genetics Initiative, the Andrew Sabin Research Fund and a Cancer Center Support Grant/Core Grant (P30 CA008748). NAROD: 1R01 CA149429-01. NCI: the Intramural

19

Research Program of the US National Cancer Institute, NIH, and by support services contracts NO2-CP-11019-50, N02-CP-21013-63 and N02-CP-65504 with Westat, Inc, Rockville, MD. NICCC: Clalit Health Services in Israel, the Israel Cancer Association and the Breast Cancer Research Foundation (BCRF), NY. NNPIO: the Russian Federation for Basic Research (grants 15-04-01744, 16-54-00055 and 17-54-12007). NRG Oncology: U10 CA180868, NRG SDMC grant U10 CA180822, NRG Administrative Office and the NRG Tissue Bank (CA 27469), the NRG Statistical and Data Center (CA 37517) and the Intramural Research Program, NCI. OSUCCG: Ohio State University Comprehensive Cancer Center. PBCS: Italian Association of Cancer Research (AIRC) [IG 2013 N.14477] and Tuscany Institute for Tumors (ITT) grant 2014-2015-2016. SEABASS: Ministry of Science, Technology and Innovation, Ministry of Higher Education (UM.C/HlR/MOHE/06) and Cancer Research Initiatives Foundation. SMC: the Israeli Cancer Association. SWE-BRCA: the Swedish Cancer Society. UCHICAGO: NCI Specialized Program of Research Excellence (SPORE) in Breast Cancer (CA125183), R01 CA142996, 1U01CA161032, American Cancer Society (MRSG-13-063-01-TBG, CRP-10-119-01-CCE), Breast Cancer Research Foundation, Susan G. Komen Foundation (SAC110026), and Ralph and Marion Falk Medical Research Trust, the Entertainment Industry Fund National Women's Cancer Research Alliance. UCLA: Jonsson Comprehensive Cancer Center Foundation; Breast Cancer Research Foundation. UCSF: UCSF Cancer Risk Program and Helen Diller Family Comprehensive Cancer Center. UKFOCR: Cancer Researc h UK. UPENN: National Institutes of Health (NIH) (R01-CA102776 and R01-CA083855; Breast Cancer Research Foundation; Susan G. Komen Foundation for the cure, Basser Research Center for BRCA. UPITT/MWH: Hackers for Hope Pittsburgh. VFCTG: Victorian Cancer Agency, Cancer Australia, National Breast Cancer Foundation. WCP: Dr Karlan is funded by the American Cancer Society Early Detection Professorship (SIOP-06-258-01-COUN) and the National Center for Advancing Translational Sciences (NCATS), Grant UL1TR000124. J Mitchell was supported by NIH grant F32CA162847.

All the families and clinicians who contribute to the studies; Sue Healey, in particular taking on the task of mutation classification with the late Olga Sinilnikova; Maggie Angelakos, Judi Maskiell, Gillian Dite, Helen Tsimiklis; members and participants in the New York site of the Breast Cancer Family Registry; members and participants in the Ontario Familial Breast Cancer Registry; Vilius Rudaitis and Laimonas Griškevičius; Drs Janis Eglitis, Anna Krilova and Aivars Stengrevics; Yuan Chun Ding and Linda Steele for their work in participant enrollment and biospecimen and data management; Bent Ejlertsen and Anne-Marie Gerdes for the recruitment and genetic counseling of participants; Alicia Barroso, Rosario Alonso and Guillermo Pita; Manoukian Siranoush, Bernard Peissel, Cristina Zanzottera, Milena Mariani, Daniela Zaffaroni, Bernardo Bonanni, Monica Barile, Irene Feroce, Mariarosaria Calvello, Alessandra Viel, Riccardo Dolcetti, Giuseppe Giannini, Laura Papi, Gabriele Lorenzo Capone, Liliana Varesco, Viviana Gismondi, Maria Grazia Tibiletti, Daniela Furlan, Antonella Savarese, Aline Martayan, Stefania Tommasi, Brunella Pilato; the personnel of the Cogentech Cancer Genetic Test Laboratory, Milan, Italy. Ms. JoEllen Weaver and Dr. Betsy Bove; Marta Santamariña, Ana Blanco, Miguel Aguado, Uxía Esperón and Belinda Rodríguez; IFE - Leipzig Research Centre for Civilization Diseases (Markus Loeffler, Joachim Thiery, Matthias Nüchter, Ronny Baber); We thank all participants, clinicians, family doctors, researchers, and technicians

20

for their contributions and commitment to the DKFZ study and the collaborating groups in Lahore, Pakistan (Muhammad U. Rashid, Noor Muhammad, Sidra Gull, Seerat Bajwa, Faiz Ali Khan, Humaira Naeemi, Saima Faisal, Asif Loya, Mohammed Aasim Yusuf) and Bogota, Colombia (Diana Torres, Ignacio Briceno, Fabian Gil). Genetic Modifiers of Cancer Risk in BRCA1/2 Mutation Carriers (GEMO) study is a study from the National Cancer Genetics Network UNICANCER Genetic Group, France. We wish to pay a tribute to Olga M. Sinilnikova, who with Dominique Stoppa-Lyonnet initiated and coordinated GEMO until she sadly passed away on the 30th June 2014. The team in Lyon (Olga Sinilnikova, Mélanie Léoné, Laure Barjhoux, Carole Verny-Pierre, Sylvie Mazoyer, Francesca Damiola, Valérie Sornin) managed the GEMO samples until the biological resource centre was transferred to Paris in December 2015 (Noura Mebirouk, Fabienne Lesueur, Dominique Stoppa-Lyonnet). We want to thank all the GEMO collaborating groups for their contribution to this study: Coordinating Centre, Service de Génétique, Institut Curie, Paris, France: Muriel Belotti, Ophélie Bertrand, Anne-Marie Birot, Bruno Buecher, Sandrine Caputo, Anaïs Dupré, Emmanuelle Fourme, Marion Gauthier-Villars, Lisa Golmard, Claude Houdayer, Marine Le Mentec, Virginie Moncoutier, Antoine de Pauw, Claire Saule, Dominique Stoppa-Lyonnet, and Inserm U900, Institut Curie, Paris, France: Fabienne Lesueur, Noura Mebirouk.Contributing Centres : Unité Mixte de Génétique Constitutionnelle des Cancers Fréquents, Hospices Civils de Lyon - Centre Léon Bérard, Lyon, France: Nadia Boutry-Kryza, Alain Calender, Sophie Giraud, Mélanie Léone. Institut Gustave Roussy, Villejuif, France: Brigitte Bressac-de-Paillerets, Olivier Caron, Marine Guillaud-Bataille. Centre Jean Perrin, Clermont–Ferrand, France: Yves-Jean Bignon, Nancy Uhrhammer. Centre Léon Bérard, Lyon, France: Valérie Bonadona, Christine Lasset. Centre François Baclesse, Caen, France: Pascaline Berthet, Laurent Castera, Dominique Vaur. Institut Paoli Calmettes, Marseille, France: Violaine Bourdon, Catherine Noguès, Tetsuro Noguchi, Cornel Popovici, Audrey Remenieras, Hagay Sobol. CHU Arnaud-de-Villeneuve, Montpellier, France: Isabelle Coupier, Pascal Pujol. Centre Oscar Lambret, Lille, France: Claude Adenis, Aurélie Dumont, Françoise Révillion. Centre Paul Strauss, Strasbourg, France: Danièle Muller. Institut Bergonié, Bordeaux, France: Emmanuelle Barouk-Simonet, Françoise Bonnet, Virginie Bubien, Michel Longy, Nicolas Sevenet, Institut Claudius Regaud, Toulouse, France: Laurence Gladieff, Rosine Guimbaud, Viviane Feillel, Christine Toulas. CHU Grenoble, France: Hélène Dreyfus, Christine Dominique Leroux, Magalie Peysselon, Rebischung. CHU Dijon, France: Amandine Baurand, Geoffrey Bertolone, Fanny Coron, Laurence Faivre, Caroline Jacquot, Sarab Lizard. CHU St-Etienne, France: Caroline Kientz, Marine Lebrun, Fabienne Prieur. Hôtel Dieu Centre Hospitalier, Chambéry, France: Sandra Fert Ferrer. Centre Antoine Lacassagne, Nice, France: Véronique Mari. CHU Limoges, France: Laurence Vénat-Bouvet. CHU Nantes, France: Stéphane Bézieau, Capucine Delnatte. CHU Bretonneau, Tours and Centre Hospitalier de Bourges France: Isabelle Mortemousque. Groupe Hospitalier Pitié-Salpétrière, Paris, France: Chrystelle Colas, Florence Coulet, Florent Soubrier, Mathilde Warcoin. CHU Vandoeuvre-les-Nancy, France: Myriam Bronner, Johanna Sokolowska. CHU Besançon, France: Marie-Agnès Collonge-Rame, Alexandre Damette. CHU Poitiers, Centre Hospitalier d’Angoulême and Centre Hospitalier de Niort, France: Paul Gesta. Centre Hospitalier de La Rochelle : Hakima Lallaoui. CHU Nîmes Carémeau, France : Jean Chiesa. CHI Poissy, France: Denise Molina-Gomes. CHU Angers, France : Olivier Ingster; Ilse Coene en Brecht Crombez; Ilse Coene and Brecht

21

Crombez; Alicia Tosar and Paula Diaque; Drs Sofia Khan, Taru A. Muranen, Carl Blomqvist, Irja Erkkilä and Virpi Palola; The Hereditary Breast and Ovarian Cancer Research Group Netherlands (HEBON) consists of the following Collaborating Centers: Coordinating center: Netherlands Cancer Institute, Amsterdam, NL: M.A. Rookus, F.B.L. Hogervorst, F.E. van Leeuwen, S. Verhoef, M.K. Schmidt, N.S. Russell, D.J. Jenner; Erasmus Medical Center, Rotterdam, NL: J.M. Collée, A.M.W. van den Ouweland, M.J. Hooning, C. Seynaeve, C.H.M. van Deurzen, I.M. Obdeijn; Leiden University Medical Center, NL: C.J. van Asperen, J.T. Wijnen, R.A.E.M. Tollenaar, P. Devilee, T.C.T.E.F. van Cronenburg; Radboud University Nijmegen Medical Center, NL: C.M. Kets, A.R. Mensenkamp; University Medical Center Utrecht, NL: M.G.E.M. Ausems, R.B. van der Luijt, C.C. van der Pol; Amsterdam Medical Center, NL: C.M. Aalfs, T.A.M. van Os; VU University Medical Center, Amsterdam, NL: J.J.P. Gille, Q. Waisfisz, H.E.J. Meijers-Heijboer; University Hospital Maastricht, NL: E.B. Gómez-Garcia, M.J. Blok; University Medical Center Groningen, NL: J.C. Oosterwijk, A.H. van der Hout, M.J. Mourits, G.H. de Bock; The Netherlands Foundation for the detection of hereditary tumours, Leiden, NL: H.F. Vasen; The Netherlands Comprehensive Cancer Organization (IKNL): S. Siesling, J.Verloop; The Dutch Pathology Registry (PALGA): L.I.H. Overbeek; Hong Kong Sanatorium and Hospital; the Hungarian Breast and Ovarian Cancer Study Group members (Janos Papp, Aniko Bozsik, Timea Pocza, Zoltan Matrai, Miklos Kasler, Judit Franko, Maria Balogh, Gabriella Domokos, Judit Ferenczi, Department of Molecular Genetics, National Institute of Oncology, Budapest, Hungary) and the clinicians and patients for their contributions to this study; the Oncogenetics Group (VHIO) and the High Risk and Cancer Prevention Unit of the University Hospital Vall d’Hebron, and the Cellex Foundation for providing research facilities and equipment; the ICO Hereditary Cancer Program team led by Dr. Gabriel Capella; the ICO Hereditary Cancer Program team led by Dr. Gabriel Capella; Dr Martine Dumont for sample management and skillful assistance; Ana Peixoto, Catarina Santos and Pedro Pinto; members of the Center of Molecular Diagnosis, Oncogenetics Department and Molecular Oncology Research Center of Barretos Cancer Hospital; Heather Thorne, Eveline Niedermayr, all the kConFab research nurses and staff, the heads and staff of the Family Cancer Clinics, and the Clinical Follow Up Study (which has received funding from the NHMRC, the National Breast Cancer Foundation, Cancer Australia, and the National Institute of Health (USA)) for their contributions to this resource, and the many families who contribute to kConFab; the KOBRA Study Group; Csilla Szabo (National Human Genome Research Institute, National Institutes of Health, Bethesda, MD, USA); Eva Machackova (Department of Cancer Epidemiology and Genetics, Masaryk Memorial Cancer Institute and MF MU, Brno, Czech Republic); and Michal Zikan, Petr Pohlreich and Zdenek Kleibl (Oncogynecologic Center and Department of Biochemistry and Experimental Oncology, First Faculty of Medicine, Charles University, Prague, Czech Republic); Anne Lincoln, Lauren Jacobs; the NICCC National Familial Cancer Consultation Service team led by Sara Dishon, the lab team led by Dr. Flavio Lejbkowicz, and the research field operations team led by Dr. Mila Pinchev; the investigators of the Australia New Zealand NRG Oncology group; members and participants in the Ontario Cancer Genetics Network; Leigha Senter, Kevin Sweet, Caroline Craven, Julia Cooper, and Michelle O'Conor; Yip Cheng Har, Nur Aishah Mohd Taib, Phuah Sze Yee, Norhashimah Hassan and all the research nurses, research assistants and doctors involved in the MyBrCa

22

Study for assistance in patient recruitment, data collection and sample preparation, Philip Iau, Sng Jen-Hwei and Sharifah Nor Akmal for contributing samples from the Singapore Breast Cancer Study and the HUKM-HKL Study respectively; the Meirav Comprehensive breast cancer center team at the Sheba Medical Center; Håkan Olsson, Helena Jernström, Karin Henriksson, Katja Harbst, Maria Soller, Ulf Kristoffersson; from Gothenburg Sahlgrenska University Hospital: Anna Öfverholm, Margareta Nordling, Per Karlsson, Zakaria Einbeigi; from Stockholm and Karolinska University Hospital: Anna von Wachenfeldt, Annelie Liljegren, Annika Lindblom, Brita Arver, Gisela Barbany Bustinza; from Umeå University Hospital: Beatrice Melin, Christina Edwinsdotter Ardnor, Monica Emanuelsson; from Uppsala University: Hans Ehrencrona, Maritta Hellström Pigg, Richard Rosenquist; from Linköping University Hospital: Marie Stenmark-Askmalm, Sigrun Liedgren; Cecilia Zvocec, Qun Niu; Joyce Seldon and Lorna Kwan; Dr. Robert Nussbaum, Beth Crawford, Kate Loranger, Julie Mak, Nicola Stewart, Robin Lee, Amie Blanco and Peggy Conrad and Salina Chan; Simon Gayther, Paul Pharoah, Carole Pye, Patricia Harrington and Eva Wozniak; Geoffrey Lindeman, Marion Harris, Martin Delatycki, Sarah Sawyer, Rebecca Driessen, and Ella Thompson for performing all DNA amplification.

Conflicts of Interest

The authors declare no conflicts of interest.

References

1. McPherson K, Steel CM, Dixon JM. ABC of breast diseases. Breast cancer-epidemiology, risk factors, and genetics. BMJ 2000;321(7261):624-8.2. Miki Y, Swensen J, Shattuck-Eidens D, et al. A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science 1994;266(5182):66-71.3. Wooster R, Bignell G, Lancaster J, et al. Identification of the breast cancer susceptibility gene BRCA2. Nature 1995;378(6559):789-92.4. Milne RL, Antoniou AC. Genetic modifiers of cancer risk for BRCA1 and BRCA2 mutation carriers. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO 2011;22 Suppl 1:i11-7.5. Friebel TM, Domchek SM, Rebbeck TR. Modifiers of cancer risk in BRCA1 and BRCA2 mutation carriers: systematic review and meta-analysis. Journal of the National Cancer Institute 2014;106(6):dju091.6. Antoniou AC, Spurdle AB, Sinilnikova OM, et al. Common breast cancer-predisposition alleles are associated with breast cancer risk in BRCA1 and BRCA2 mutation carriers. Am J Hum Genet 2008;82(4):937-48.7. Kotsopoulos J, Lubinski J, Lynch HT, et al. Oophorectomy after menopause and the risk of breast cancer in BRCA1 and BRCA2 mutation carriers. Cancer Epidemiol Biomarkers Prev 2012;21(7):1089-96.8. Narod SA. Modifiers of risk of hereditary breast cancer. Oncogene 2006;25(43):5832-6.9. Rebbeck TR, Mitra N, Wan F, et al. Association of type and location of BRCA1 and BRCA2 mutations with risk of breast and ovarian cancer. JAMA 2015;313(13):1347-61.

23

10. Couch FJ, Wang X, McGuffog L, et al. Genome-wide association study in BRCA1 mutation carriers identifies novel loci associated with breast and ovarian cancer risk. PLoS Genet 2013;9(3):e1003212.11. Gaudet MM, Kuchenbaecker KB, Vijai J, et al. Identification of a BRCA2-specific modifier locus at 6p24 related to breast cancer risk. PLoS Genet 2013;9(3):e1003173.12. van den Brandt PA, Spiegelman D, Yaun SS, et al. Pooled analysis of prospective cohort studies on height, weight, and breast cancer risk. Am J Epidemiol 2000;152(6):514-27.13. Trentham-Dietz A, Newcomb PA, Storer BE, et al. Body size and risk of breast cancer. Am J Epidemiol 1997;145(11):1011-9.14. Zhang B, Shu XO, Delahanty RJ, et al. Height and Breast Cancer Risk: Evidence From Prospective Studies and Mendelian Randomization. Journal of the National Cancer Institute 2015;107(11).15. Renehan AG, Tyson M, Egger M, et al. Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet 2008;371(9612):569-78.16. Locke AE, Kahali B, Berndt SI, et al. Genetic studies of body mass index yield new insights for obesity biology. Nature 2015;518(7538):197-206.17. Wood AR, Esko T, Yang J, et al. Defining the role of common variation in the genomic and biological architecture of adult human height. Nat Genet 2014;46(11):1173-86.18. Lango Allen H, Estrada K, Lettre G, et al. Hundreds of variants clustered in genomic loci and biological pathways affect human height. Nature 2010;467(7317):832-8.19. Palmer TM, Sterne JA, Harbord RM, et al. Instrumental variable estimation of causal risk ratios and causal odds ratios in Mendelian randomization analyses. Am J Epidemiol 2011;173(12):1392-403.20. Khankari NK, Shu XO, Wen W, et al. Association between Adult Height and Risk of Colorectal, Lung, and Prostate Cancer: Results from Meta-analyses of Prospective Studies and Mendelian Randomization Analyses. PLoS Med 2016;13(9):e1002118.21. VanderWeele TJ, Tchetgen Tchetgen EJ, Cornelis M, et al. Methodological challenges in mendelian randomization. Epidemiology 2014;25(3):427-35.22. Guo Y, Warren Andersen S, Shu XO, et al. Genetically Predicted Body Mass Index and Breast Cancer Risk: Mendelian Randomization Analyses of Data from 145,000 Women of European Descent. PLoS Med 2016;13(8):e1002105.23. Antoniou AC, Goldgar DE, Andrieu N, et al. A weighted cohort approach for analysing factors modifying disease risks in carriers of high-risk susceptibility genes. Genet Epidemiol 2005;29(1):1-11.24. Antoniou A, Pharoah PD, Narod S, et al. Average risks of breast and ovarian cancer associated with BRCA1 or BRCA2 mutations detected in case Series unselected for family history: a combined analysis of 22 studies. Am J Hum Genet 2003;72(5):1117-30.25. Thomas DC, Lawlor DA, Thompson JR. Re: Estimation of bias in nongenetic observational studies using "Mendelian triangulation" by Bautista et al. Ann Epidemiol 2007;17(7):511-3.26. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol 2013;37(7):658-65.27. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol 2015;44(2):512-25.

24

28. Terza JV, Basu A, Rathouz PJ. Two-stage residual inclusion estimation: addressing endogeneity in health econometric modeling. J Health Econ 2008;27(3):531-43.29. Manders P, Pijpe A, Hooning MJ, et al. Body weight and risk of breast cancer in BRCA1/2 mutation carriers. Breast Cancer Res Treat 2011;126(1):193-202.30. Kotsopoulos J, Olopado OI, Ghadirian P, et al. Changes in body weight and the risk of breast cancer in BRCA1 and BRCA2 mutation carriers. Breast Cancer Res 2005;7(5):R833-43.31. Gunnell D, Okasha M, Smith GD, et al. Height, leg length, and cancer risk: a systematic review. Epidemiol Rev 2001;23(2):313-42.32. Okasha M, McCarron P, Gunnell D, et al. Exposures in childhood, adolescence and early adulthood and breast cancer risk: a systematic review of the literature. Breast Cancer Res Treat 2003;78(2):223-76.33. Pollak M. The insulin and insulin-like growth factor receptor family in neoplasia: an update. Nat Rev Cancer 2012;12(3):159-69.34. Hankinson SE, Willett WC, Colditz GA, et al. Circulating concentrations of insulin-like growth factor-I and risk of breast cancer. Lancet 1998;351(9113):1393-6.35. Stefan N, Haring HU, Hu FB, et al. Divergent associations of height with cardiometabolic disease and cancer: epidemiology, pathophysiology, and global implications. Lancet Diabetes Endocrinol 2016;4(5):457-67.36. Qian F, Feng Y, Zheng Y, et al. Genetic variants in microRNA and microRNA biogenesis pathway genes and breast cancer risk among women of African ancestry. Hum Genet 2016;135(10):1145-59.37. Lettre G, Jackson AU, Gieger C, et al. Identification of ten loci associated with height highlights new biological pathways in human growth. Nat Genet 2008;40(5):584-91.38. Calle EE, Kaaks R. Overweight, obesity and cancer: epidemiological evidence and proposed mechanisms. Nat Rev Cancer 2004;4(8):579-91.39. Michailidou K, Hall P, Gonzalez-Neira A, et al. Large-scale genotyping identifies 41 new loci associated with breast cancer risk. Nat Genet 2013;45(4):353-61, 361e1-2.40. Couch FJ, Kuchenbaecker KB, Michailidou K, et al. Identification of four novel susceptibility loci for oestrogen receptor negative breast cancer. Nat Commun 2016;7:11375.41. Tan A, Dang Y, Chen G, et al. Overexpression of the fat mass and obesity associated gene (FTO) in breast cancer and its clinical implications. Int J Clin Exp Pathol 2015;8(10):13405-10.42. Zeng X, Ban Z, Cao J, et al. Association of FTO Mutations with Risk and Survival of Breast Cancer in a Chinese Population. Dis Markers 2015;2015:101032.43. Hernandez-Caballero ME, Sierra-Ramirez JA. Single nucleotide polymorphisms of the FTO gene and cancer risk: an overview. Mol Biol Rep 2015;42(3):699-704.

25

Table 1. Baseline characteristics of participants in the CIMBA consortium with genotype information

VariableBRCA1 carriers

N = 14,676BRCA2 carriers

N = 7,912

Case participants, n 7,360 4,091Year of birth, median (IQR) 1956 (1948 - 1964) 1954 (1945 - 1961)Age at diagnosis, y (mean ± SD) 41.1 ± 9.3 44.2 ± 10.0

Control participants, n 7,316 3,821Year of birth, median (IQR) 1963 (1953 - 1972) 1962 (1950 - 1971)Age at censoring, y (mean ± SD) 42.0 ± 12.5 44.0 ± 13.5

Ethnicity, n (%)Caucasian, not otherwise specified 13,435 (91.5) 7,126 (90.1)Ashkenazi Jewish 1,241 (8.5) 786 (9.9)

Height in cm, n 4,502 3,155

Mean ± SD 164.8 ± 6.8 164.5 ± 6.9

Weight at date of questionnaire in kg, n 4,436 3,133 Mean ± SD 68.1 ± 13.9 69.0 ± 14.5

Body mass index at baseline1 in kg/m2 , n 4,401 3,115 Mean ± SD 25.1 ± 5.1 25.6 ± 5.2Weight in early adulthood in kg, n 3,152 2,296 Mean ± SD 57.6 ± 9.2 57.9 ± 9.6Body mass index in early adulthood in kg/m2

, n 3,134 2,283 Mean ± SD 21.2 ± 3.3 21.4 ± 3.3Age at menarche in years, n 4,425 3,034 Mean ± SD 13.0 ± 1.5 13.0 ± 1.5Parous, n (%)

Yes 3,914 (77.8) 2,681(79.7)No 1,117 (22.2) 682 (20.3)

Age at first live birth in years, n 3,711 2,579 Mean ± SD 25.3 ± 4.9 25.3 ± 4.9Menopausal status, n (%)

Premenopausal 2,330 (47.4) 1,610 (46.4)Postmenopausal 2,588 (52.6) 1,858 (53.6)

Age at menopause, y (mean ± SD) 44.7 ± 6.0 45.6 ± 6.0

Table 2. Association of height and breast cancer risk using observed height, among 7,657 participants

26

N/events Hazard Ratio (95% CI) P-value

Per 10 cm increase in observed heightAll participants (confounding adjustment sequentially) Unadjusted 7657/3653 1.14 (1.06 - 1.22) 0.0002

Adjusted for principal components 7657/3653 1.15 (1.07 - 1.23) 0.0002

Additionally adjusted for country 7657/3653 1.17 (1.09 - 1.26)<0.000

1Additionally adjusted for birth cohort 7657/3653 1.09 (1.01 - 1.17) 0.021Additionally adjusted for mutation status 7657/3653 1.09 (1.02 - 1.17) 0.012Additionally adjusted for menopausal status 7657/3653 1.09 (1.02 - 1.17) 0.016Additionally adjusted for parity and age at

menarche 7090/3383 1.10 (1.02 - 1.18) 0.014

By mutation status*

BRCA1 carrier 4502/2154 1.06 (0.97 - 1.16) 0.19BRCA2 carrier 3155/1499 1.17 (1.04 - 1.31) 0.007Pinteraction 0.18

By menopausal status**

Premenopausal 7657/2197 1.10 (1.01 - 1.20) 0.025Postmenopausal 3076/1402 1.07 (0.95 - 1.19) 0.26Pinteraction 0.64

Notes:Notes:* Adjusted for prinicpal components, birth cohort, country of enrollment, and menopausal status** Adjusted for principal components, mutation status, birth cohort, and country of enrollment.

27

Table 3. Associations of the height genetic score (height-GS) with height and traditional breast cancer risk factors

VariableNumber of participants

Summary effect

Standard error P-value

% variation explained

Measured height, cm

All participants 7,657 1.012 0.029 7.0 x 10-241 13.4

BRCA1 carriers 4,502 1.025 0.038 3.8 x 10-149 14.0

BRCA2 carriers 3,155 0.996 0.047 2.5 x 10-94 12.6

Case participants 3,653 1.028 0.041 2.6 x 10-128 14.7

Control participants 4,004 1.000 0.042 1.1 x 10-116 12.3

Traditional risk factors

BMI, kg/m2 7,516 -0.010 0.024 0.66

Weight, kg 7,569 0.813 0.065 2.7 x 10-35

Age at baseline, y 8,578 0.123 0.036 7.3 x 10-4

Age at menarche, y 7,459 0.028 0.007 9.3 x 10-5

Parous, yes vs no 8,394 -0.010 0.011 0.35

Age at first live birth, y 6,290 -0.014 0.025 0.58

Menopausal status, pre vs post 8,386 0.011 0.009 0.20

Age at menopause, y 4,336 -0.082 0.037 0.03

Notes:Regression coefficient is presented for continuous variables and natural log-scale odds ratio for binary variables, per unit increase of the weighted height genetic score

28

Table 4. Association of height and breast cancer risk among 22,588 participants in CIMBA, per 10 cm increase in height

Breast cancer group N/eventsHazard Ratio (95%

CI) P-valueHeterogeneity

(I2)

Height-GS*All participants (confounding adjustment sequentially)

Unadjusted 22588/11451 1.11 (1.00 - 1.23) 0.05Adjusted for principal components 22588/11451 1.04 (0.93 - 1.17) 0.48Additionally adjusted for country 22588/11451 1.03 (0.92 - 1.16) 0.57Additionally adjusted for birth cohort 22588/11451 1.04 (0.92 - 1.16) 0.56Additionally adjusted for mutation status 22588/11451 1.04 (0.93 - 1.17) 0.45Additionally adjusted for menopausal status 22588/11451 1.04 (0.93 - 1.17) 0.47

By mutation status**BRCA1 carrier 14676/7360 1.03 (0.91 - 1.18) 0.62BRCA2 carrier 7912/4091 1.07 (0.87 - 1.32) 0.50Pinteraction 0.95

By menopausal status†Premenopausal 22588/7410 1.09 (0.96 - 1.24) 0.20Postmenopausal 8459/3926 0.95 (0.79 - 1.13) 0.55Pinteraction 0.18

Meta-analysis method ‡ All participants 22588/11451 1.05 (0.93 - 1.19) 0.42 17.0%

BRCA1 carrier 14676/7360 1.04 (0.90 - 1.20) 0.57 11.8%BRCA2 carrier 7912/4091 1.09 (0.87 - 1.36) 0.45 6.6%Pinteraction 0.75

Two-stage residual inclusion methodAll participants 7657/3653 1.09 (0.93 - 1.27) 0.27

BRCA1 carrier 4502/2154 1.16 (0.96 - 1.40) 0.13BRCA2 carrier 3155/1499 1.05 (0.80 - 1.37) 0.74

Notes:* Height genetic score combining 586 height-associated single nucleoid polymorphsims**Adjusted for principal components, birth cohort, country of enrollment, and menopausal status† Adjusted for principal components, mutation status, birth cohort, and country of enrollment.‡ Hazard ratios were calculated using inverse-variance meta-analysis and re-scaled to the corresponding units by calculating the height measurements per z-score among controls. Effect estimates for breast cancer for each SNP were calculated from weighted Cox model adjusting for birth cohort, country of enrollment, menopausal status, and mutation status.

29

Table 5. Association of BMI and breast cancer risk using observed BMI

Breast cancer group N/events Hazard Ratio (95% CI) P-value

Per 5 kg/m2 increase in BMI at date of questionnaireAll participants (confounding adjustment sequentially)

Unadjusted 7516/3594 0.92 (0.88 - 0.97) 0.0004Adjusted for principal components 7516/3594 0.93 (0.89 - 0.97) 0.0007Additionally adjusted for country 7516/3594 0.92 (0.88 - 0.96) 0.0003Additionally adjusted for birth cohort 7516/3594 0.94 (0.90 - 0.98) 0.003Additionally adjusted for mutation status 7516/3594 0.95 (0.91 - 0.99) 0.014Additionally adjusted for menopausal status 7516/3594 0.94 (0.90 - 0.98) 0.007Additionally adjusted for parity and age at menarche 6964/3331 0.93 (0.89 - 0.98) 0.003

By mutation status*BRCA1 carrier 4401/2114 0.96 (0.91 - 1.01) 0.11BRCA2 carrier 3115/1480 0.90 (0.84 - 0.97) 0.003Pinteraction 0.26

By menopausal status**Premenopausal 7516/2153 0.92 (0.87 - 0.97) 0.001Postmenopausal 3029/1389 0.97 (0.91 - 1.04) 0.40Pinteraction 0.14

Per 5 kg/m2 increase in BMI in young adulthoodAll participants (confounding adjustment sequentially)

Unadjusted 5417/2520 0.83 (0.76 - 0.91) 3.1E-05Adjusted for principal components 5417/2520 0.83 (0.76 - 0.91) 5.4E-05Additionally adjusted for country 5417/2520 0.81 (0.74 - 0.88) 2.8E-06Additionally adjusted for birth cohort 5417/2520 0.81 (0.74 - 0.88) 1.8E-06Additionally adjusted for mutation status 5417/2520 0.83 (0.76 - 0.90) 1.7E-05Additionally adjusted for menopausal status 5417/2520 0.83 (0.76 - 0.90) 2.1E-05Additionally adjusted for parity and age at menarche 5210/2436 0.82 (0.75 - 0.90) 2.7E-05

By mutation status*BRCA1 carrier 3134/1462 0.87 (0.78 - 0.97) 0.013BRCA2 carrier 2283/1058 0.74 (0.63 - 0.85) 4.5E-05Pinteraction 0.06

By menopausal status**Premenopausal 5417/1519 0.85 (0.78 - 0.94) 0.002Postmenopausal 2181/977 0.79 (0.69 - 0.91) 0.001Pinteraction 0.35

Notes:* Adjusted for principal components, birth cohort, country of enrollment, and menopausal status** Adjusted for principal components, mutation status, birth cohort, and country of enrollment.

30

Table 6. Associations of the body mass index genetic score (BMI-GS) with BMI and traditional breast cancer risk factors

VariableNumber of

participants Summary effect* Standard error P-value

% variation explained

Observed BMI at date of questionnaire, kg/m2

All participants 7,516 0.832 0.059 2.8 x 10-44 2.6BRCA1 carrier 4,401 0.786 0.076 9.9 x 10-25 2.4BRCA2 carrier 3,115 0.903 0.094 2.1 x 10-21 2.9

Case participants 3,594 0.651 0.083 6.9 x 10-15 1.7Control participants 3,922 1.000 0.084 3.3 x 10-32 3.5Premenopausal control participants 2,246 0.941 0.110 2.6 x 10-17 3.1Postmenopausal control participants 1,441 1.113 0.139 3.0 x 10-15 4.2

Observed BMI in young adulthood, kg/m2

All participants 5,417 0.794 0.083 2.3 x 10-21 1.6BRCA1 carrier 3,134 0.770 0.108 1.2 x 10-12 1.6BRCA2 carrier 2,283 0.830 0.131 2.8 x 10-10 1.7

Case participants 2,520 0.540 0.110 1.0 x 10-6 0.9Control participants 2,897 1.000 0.122 3.3 x 10-16 2.3Premenopausal control participants 1,589 0.995 0.170 5.3 x 10-9 2.1Postmenopausal control participants 1,100 1.044 0.193 8.2 x 10-8 2.6

Traditional risk factorsHeight 7,657 0.319 0.079 5.4 x 10-5

Age at baseline, y 8,578 -0.020 -0.020 0.79Age at menarche, y 7,459 -0.090 0.018 5.0 x 10-7

Parous, yes vs no 8,394 0.024 0.027 0.38Age at first live birth, y 6,290 -0.113 0.063 0.07Menopausal status, pre vs post 8,386 -0.019 0.022 0.39Age at menopause, y 4,336 -0.172 0.093 0.06

Notes:* Regression coefficient is presented for continuous variables and natural log-scale odds ratio for binary variables, per unit increase of the weighted BMI genetic score

31

Table 7. Association of BMI and breast cancer risk among 22,588 participants in CIMBA, per 5 kg/m2 increase in BMI

Breast cancer group N/events HR (95% CI)P-

valueHeterogeneity

(I2)

BMI-GS*All participants (confounding adjustment sequentially)

Unadjusted 22588/11451 0.93 (0.81 - 1.05) 0.24Adjusted for principal components 22588/11451 0.89 (0.78 - 1.01) 0.071Additionally adjusted for country 22588/11451 0.90 (0.79 - 1.03) 0.13Additionally adjusted for birth cohort 22588/11451 0.88 (0.77 - 0.999) 0.049Additionally adjusted for mutation status 22588/11451 0.88 (0.78 - 0.99) 0.039Additionally adjusted for menopausal status 22588/11451 0.87 (0.76 - 0.98) 0.023

By mutation status**BRCA1 carrier 14676/7360 0.88 (0.76 - 1.02) 0.09BRCA2 carrier 7912/4091 0.83 (0.65 - 1.05) 0.11Pinteraction 0.63

By menopausal status†Premenopausal 22588/7410 0.84 (0.73 - 0.98) 0.023Postmenopausal 8459/3926 0.89 (0.72 - 1.09) 0.26Pinteraction 0.68

Meta-analysis method ‡ All participants 22588/11451 0.87 (0.76 - 0.98) 0.027 3.5%

BRCA1 carrier 14676/7360 0.88 (0.76 - 1.03) 0.10 15.7%BRCA2 carrier 7912/4091 0.82 (0.65 - 1.04) 0.10 0.0%Pinteraction 0.63

Two-stage residual inclusion methodAll participants 7516/3594 0.86 (0.70 - 1.07) 0.18

BRCA1 carrier 4401/2114 0.93 (0.69 - 1.23) 0.61BRCA2 carrier 3115/1480 0.82 (0.61 - 1.12) 0.23

Notes:* BMI-GS was constructed by combining 93 BMI-associated SNPs.** Adjusted for principal components, birth cohort, country of enrollment, and menopausal status† Adjusted for mutation status, birth cohort, and country of enrollment.

‡ Hazard ratios were calculated using inverse-variance meta-analysis and re-scaled to the corresponding units by calculating the BMI measurements per z-score among controls. Effect estimates for breast cancer for each SNP were calculated from weighted Cox model adjusting for birth cohort, country of enrollment, menopausal status, and mutation status.

32

33

Table 8. Height or BMI SNPs significantly associated (P < 0.05) with breast cancer risk in CIMBA

Rsid Chromosome Position Nearest gene

Reference allele in CIMBA

Effect allele in CIMBA

Effect allele frequency in CIMBA

Imputation quality*

Association with breast cancer in CIMBALog hazard

ratio** Standard error P-value

Height

rs10744956 15 51269629 AP4E1 A G 0.80 0.98 0.096 0.023 0.00003

rs7740107 6 130374461 L3MBTL3 T A 0.75 1 0.081 0.021 0.0001

rs10995319 10 52762887 PRKG1 T C 0.23 0.96 0.075 0.022 0.0006

rs8058684 16 53515118 RBL2 G A 0.32 1 0.061 0.019 0.0013

rs11049611 12 28600244 CCDC91 C T 0.28 1 -0.065 0.020 0.0015

rs8103992 19 19665643 PBX4 A C 0.78 0.98 -0.064 0.022 0.0036

rs11618507 13 30172751 SLC7A1 G T 0.20 1 0.061 0.023 0.0074

rs11244750 10 127673877 FANK1 C T 0.35 0.83 0.055 0.021 0.0089

rs7701414 5 131585958 PDLIM4 A G 0.46 1 0.047 0.018 0.01

rs7733195 5 172994624 FAM44B G A 0.37 1 0.049 0.019 0.0107

rs2306694 12 56680636 CS A G 0.06 1 0.096 0.038 0.012

rs6435143 2 203194256 NOP5/NOP A C 0.56 0.84 0.050 0.020 0.013

rs2284746 1 17306675 MFAP2 C G 0.50 0.84 -0.049 0.020 0.013

rs1797625 3 112826415 C3orf17 A T 0.36 0.91 -0.049 0.020 0.013

rs10495098 1 218516310 TGFB2 G T 0.41 0.84 -0.051 0.021 0.013

rs301901 5 37046626 NIPBL A G 0.45 0.93 -0.046 0.019 0.015

34

rs42039 7 92244422 CDK6 C T 0.26 1 0.051 0.021 0.016

rs1576900 9 18629792 ADAMTSL1 G A 0.30 0.91 0.048 0.020 0.017

rs891088 19 7184762 INSR A G 0.27 0.83 0.052 0.022 0.020

rs273945 7 137611566 CREB3L2 A C 0.58 1 -0.042 0.018 0.021

rs2682587 19 44082429 XRCC1 C A 0.18 1 0.053 0.023 0.022

rs1257763 9 96893945 PTPDC1 A G 0.96 0.72 -0.118 0.053 0.026

rs2888877 7 92228400 CDK6 T C 0.79 0.96 -0.050 0.022 0.027

rs8042424 15 101762539 CHSY1 C T 0.24 0.82 -0.052 0.024 0.028

rs7716219 5 54955071 SLC38A9 T C 0.70 0.97 -0.044 0.020 0.029

rs7727731 5 64674446 ADAMTS6 C T 0.12 0.67 -0.076 0.035 0.029

rs16964211 15 51530495 CYP19A1 G A 0.06 0.99 -0.085 0.040 0.031

rs11880992 19 2176403 DOT1L G A 0.42 0.95 -0.040 0.019 0.033

rs2302580 4 8608634 CPZ C T 0.44 1 -0.040 0.019 0.033

rs12538407 7 23521316 IGF2BP3 A G 0.39 0.99 0.040 0.019 0.034

rs2300921 3 185651001 SFRS10 T C 0.41 0.98 0.041 0.019 0.034

rs897080 2 44774202 C2orf34 C T 0.79 0.90 -0.050 0.024 0.034

rs4357716 11 69163161 MYEOV C T 0.14 1 0.055 0.026 0.034

rs11659752 18 77222862 NFATC1 T G 0.30 0.82 -0.045 0.021 0.035

rs2166898 2 121612659 GLI2 G A 0.17 0.55 0.067 0.032 0.036

rs6020202 20 48634821 SNAI1 G A 0.24 1 0.045 0.022 0.036

35

rs3760318 17 29247715 CENTA2 G A 0.37 1 -0.039 0.019 0.037

rs6746356 2 174815898 SP3 A C 0.24 0.79 0.050 0.024 0.037

rs9428104 1 118855587 SPAG17 A G 0.75 1 0.044 0.021 0.037

rs2834442 21 35690786 KCNE2 T A 0.66 0.98 -0.040 0.019 0.038

rs1546391 3 114697457 ZBTB20 C G 0.09 0.97 -0.069 0.033 0.038

rs12926008 16 2488211 CCNF T C 0.65 0.53 0.054 0.026 0.040

rs2123731 19 4929473 UHRF1 A G 0.27 0.70 -0.049 0.024 0.043

rs2289195 2 25463483 DNMT3A G A 0.41 0.65 0.046 0.023 0.043

rs16834765 1 32371442 PTP4A2 C T 0.05 0.82 0.091 0.045 0.044

rs10958476 8 57095808 PLAG1 T C 0.20 1 0.045 0.022 0.045

rs17369123 1 172355841 DNM3 C T 0.16 0.98 0.049 0.025 0.046

rs1415701 6 130345835 L3MBTL3 G A 0.26 1 0.041 0.021 0.047

rs11152213 18 57852948 MC4R A C 0.23 0.99 -0.043 0.022 0.047

rs4802134 19 38346685 SIPA1L3 A G 0.75 1 -0.041 0.021 0.049

BMI

rs13107325 4 103188709 SLC39A8 C T 0.09 0.76 0.101 0.037 0.007

rs10182181 2 25150296 ADCY3 A G 0.45 0.64 -0.056 0.023 0.016

rs7903146 10 114758349 TCF7L2 C T 0.30 1 0.044 0.020 0.025

rs9925964 16 31129895 KAT8 A G 0.39 0.99 -0.040 0.019 0.034

36

rs4740619 9 15634326 C9orf93 T C 0.46 0.97 0.040 0.019 0.034

rs1558902 16 53803574 FTO T A 0.42 1.00 -0.038 0.018 0.037

rs2207139 6 50845490 TFAP2B A G 0.16 0.99 0.049 0.024 0.044

Notes:

* Imputation quality of 1 indicates genotyped SNPs.** Per-allele association with breast cancer was adjusted for principal components, birth cohort, menopausal status, age at menopause, country of enrollment and mutation status in weighted Cox models